public void TestSignaturesV2_3_X()
{
var intMemoryStream = new MemoryStream <int>(_spark);
StreamingQuery sq1 = intMemoryStream
.ToDF().WriteStream().QueryName("intQuery").Format("console").Start();
string id1 = sq1.Id;
var stringMemoryStream = new MemoryStream <string>(_spark);
StreamingQuery sq2 = stringMemoryStream
.ToDF().WriteStream().QueryName("stringQuery").Format("console").Start();
string id2 = sq2.Id;
StreamingQueryManager sqm = _spark.Streams();
StreamingQuery[] streamingQueries = sqm.Active().ToArray();
Assert.Equal(2, streamingQueries.Length);
Assert.IsType <StreamingQuery>(sqm.Get(id1));
Assert.IsType <StreamingQuery>(sqm.Get(id2));
sqm.ResetTerminated();
sqm.AwaitAnyTermination(10);
sq1.Stop();
sq2.Stop();
}
public void TestSignaturesV2_4_X()
{
var intMemoryStream = new MemoryStream <int>(_spark);
StreamingQuery sq = intMemoryStream
.ToDF()
.WriteStream()
.QueryName("testQuery")
.Format("console")
.Trigger(Trigger.Once())
.Start();
sq.AwaitTermination();
Assert.IsType <bool>(sq.AwaitTermination(10));
Assert.IsType <string>(sq.Name);
Assert.IsType <string>(sq.Id);
Assert.IsType <string>(sq.RunId);
Assert.IsType <bool>(sq.IsActive());
sq.Explain();
Assert.Null(sq.Exception());
sq.Stop();
}
private static void SetupExitHandlers()
{
Console.CancelKeyPress += (s, e) =>
{
Console.WriteLine($"{Environment.NewLine}Ctrl+C pressed");
Environment.Exit(0);
};
AppDomain.CurrentDomain.ProcessExit += (s, e) =>
{
Console.WriteLine($"{Environment.NewLine}Exiting");
// stop spark query
s_query.Stop();
// Allow the main thread to continue and exit...
s_waitHandle.Set();
};
}
public void TestSignaturesV3_1_X()
{
string tableName = "output_table";
WithTable(
_spark,
new string[] { tableName },
() =>
{
using var tempDirectory = new TemporaryDirectory();
var intMemoryStream = new MemoryStream <int>(_spark);
DataStreamWriter dsw = intMemoryStream
.ToDF()
.WriteStream()
.Format("parquet")
.Option("checkpointLocation", tempDirectory.Path);
StreamingQuery sq = dsw.ToTable(tableName);
sq.Stop();
});
}
public void TestForeachBatch()
{
// Temporary folder to put our test stream input.
using var srcTempDirectory = new TemporaryDirectory();
// Temporary folder to write ForeachBatch output.
using var dstTempDirectory = new TemporaryDirectory();
Func <Column, Column> outerUdf = Udf <int, int>(i => i + 100);
// id column: [0, 1, ..., 9]
WriteCsv(0, 10, Path.Combine(srcTempDirectory.Path, "input1.csv"));
DataStreamWriter dsw = _spark
.ReadStream()
.Schema("id INT")
.Csv(srcTempDirectory.Path)
.WriteStream()
.ForeachBatch((df, id) =>
{
Func <Column, Column> innerUdf = Udf <int, int>(i => i + 200);
df.Select(outerUdf(innerUdf(Col("id"))))
.Write()
.Csv(Path.Combine(dstTempDirectory.Path, id.ToString()));
});
StreamingQuery sq = dsw.Start();
// Process until all available data in the source has been processed and committed
// to the ForeachBatch sink.
sq.ProcessAllAvailable();
// Add new file to the source path. The spark stream will read any new files
// added to the source path.
// id column: [10, 11, ..., 19]
WriteCsv(10, 10, Path.Combine(srcTempDirectory.Path, "input2.csv"));
// Process until all available data in the source has been processed and committed
// to the ForeachBatch sink.
sq.ProcessAllAvailable();
sq.Stop();
// Verify folders in the destination path.
string[] csvPaths =
Directory.GetDirectories(dstTempDirectory.Path).OrderBy(s => s).ToArray();
var expectedPaths = new string[]
{
Path.Combine(dstTempDirectory.Path, "0"),
Path.Combine(dstTempDirectory.Path, "1"),
};
Assert.True(expectedPaths.SequenceEqual(csvPaths));
// Read the generated csv paths and verify contents.
DataFrame df = _spark
.Read()
.Schema("id INT")
.Csv(csvPaths[0], csvPaths[1])
.Sort("id");
IEnumerable <int> actualIds = df.Collect().Select(r => r.GetAs <int>("id"));
Assert.True(Enumerable.Range(300, 20).SequenceEqual(actualIds));
}
